• Infrared and Laser Engineering
  • Vol. 49, Issue 6, 20200010 (2020)
Feng Shi1, Tongxi Lu2, Shuning Yang1, Zhuang Miao1, Ye Yang1, Wenwen Zhang2, and Ruiqing He3、*
Author Affiliations
  • 1微光夜视技术重点实验室,陕西 西安 710065
  • 2南京理工大学 江苏省光谱成像和智能感知重点实验室,江苏 南京 210094
  • 3南京工程学院 信息与通信工程学院,江苏 南京 211167
  • show less
    DOI: 10.3788/IRLA20200010 Cite this Article
    Feng Shi, Tongxi Lu, Shuning Yang, Zhuang Miao, Ye Yang, Wenwen Zhang, Ruiqing He. Target recognition method based on single-pixel imaging system and deep learning in the noisy environment[J]. Infrared and Laser Engineering, 2020, 49(6): 20200010 Copy Citation Text show less
    Procedure of experiment
    Fig. 1. Procedure of experiment
    Classical ghost imaging system based on thermal source
    Fig. 2. Classical ghost imaging system based on thermal source
    Composition of ResNet and ResNet _V2. (a) Composition of ResNet; (b) composition of ResNet_V2
    Fig. 3. Composition of ResNet and ResNet _V2. (a) Composition of ResNet; (b) composition of ResNet_V2
    Rearrangement of non-imaged samples and pseudo-color mapping method
    Fig. 4. Rearrangement of non-imaged samples and pseudo-color mapping method
    Preprocessing (a) and inferring (b) time for the recognition methods using non-imaged and imaged sample respectively
    Fig. 5. Preprocessing (a) and inferring (b) time for the recognition methods using non-imaged and imaged sample respectively
    Recognition accuracy for the targets with different sparsity ratios
    Fig. 6. Recognition accuracy for the targets with different sparsity ratios
    ε=20% ε=30% ε=40%
    β=4.7% 99%97%87%
    β=9.7% 98%100%90%
    β=21.9% 100%91%72%
    β=39% 97%88%58%
    Table 1. Recognition results using non-imaged samples
    ε=20% ε=30% ε=40%
    β=4.7% 99%98%89%
    β=9.7% 100%94%91%
    β=21.9% 100%100%97%
    β=39% 100%100%98%
    Table 2. Recognition results using imaged samples
    Feng Shi, Tongxi Lu, Shuning Yang, Zhuang Miao, Ye Yang, Wenwen Zhang, Ruiqing He. Target recognition method based on single-pixel imaging system and deep learning in the noisy environment[J]. Infrared and Laser Engineering, 2020, 49(6): 20200010
    Download Citation